Sensitivity of ASCE-Penman–Monteith reference evapotranspiration under different climate types in Brazil

Abstract

Sensitivity analysis of reference evapotranspiration (ETo) plays a key role in the simplification and improvement of measurements of terrestrial water balance. The aim of this study was to perform sensitivity analyses of the ASCE-Penman–Monteith reference ET equation (EToPM) for different tropical and subtropical climates, where a quantitative understanding of water fluxes to the atmosphere is limited. Sensitivity coefficients were derived on a daily basis for maximum and minimum air temperature, solar radiation, vapor pressure deficit and wind speed at 2 m height using data from 44-year of actual measurements for calibration (1970–2014), for nine different climatic zones across Brazil. A multiple regression analysis was performed to estimate the relation between meteorological data and EToPM across climatic zones. Seasonal and annual average estimates were obtained by averaging daily values and spatial patterns of EToPM were obtained by interpolating meteorological data from all sampled locations. Five climate variables were used in the analysis, which revealed diverse effects on EToPM across seasons and climatic zones. In order of importance, EToPM was most sensitive to annual variation in vapor pressure deficit (VPD), wind speed (U2) and solar radiation (Rs) in all climate types. Our analysis also showed that VPD, calculated from measurements of relative humidity and temperature (T), are essential to accurately predict EToPM across tropical and subtropical climates. Due to the lack of direct meteorological measurements in many Brazilian regions, we recommend the adjustment of climate-driven hydrological fluxes predictions to the most sensitive variables, i.e., VPD, to improve the precision of reference ET losses. Our results will be useful in delineating the influence of different climatic variables in the ASCE Penman–Monteith model and in guiding new climatic modeling efforts in tropical and subtropical regions.